The guest for our July 2026 members’ meeting was Jaap de Vries, Principal Innovation Specialist at FM Global, and an adjunct professor at Brown University’s Master’s Program in Innovation Management & Entrepreneurship.
His definition of vibe entrepreneurship: AI‑mediated venture creation in which founders or small teams move from a hypothesis to a testable business artifact, primarily through natural language interaction with AI systems, and while using those artifacts to accelerate customer learnings, business learnings, and product iteration.
And vibe intrapreneurship: AI‑mediated internal venture building in which employees or cross‑functional teams use AI to create working artifacts, internal business tools, new product concepts, operating pilots, business cases, and stakeholder narrative much earlier and more cheaply than before, while still operating within enterprise constraints.
Some of the advice he shared during this session:
1. Use AI across the full innovation funnel, not just for coding
De Vries maps six phases — sense, ideate, validate, build, pilot, scale — and argues AI should be applied conversationally at each step, not only for writing code or slideware.
2. Practice “context engineering,” not just prompt engineering
He stresses that giving AI the right corporate context (strategy docs, department priorities, internal data) is crucial; otherwise you get commoditized, generic answers or outputs misaligned with strategy.
3. Systematically mine internal signals, but let humans decide what matters
De Vries recommends leveraging AI on meeting transcripts, interviews, surveys, tickets, operating data to sense opportunities and pain points, while warning about “signal flooding” and the need for human judgment to separate signal from noise.
4. Use synthetic personas/panels as a first pass, not a replacement for real customers
He shows how AI‑generated personas and virtual panels can help pre‑test decks, concepts, and packaging, but warns against:
- Over‑trusting flattering feedback (“10/10” ratings)
- Confusing this with real customer discovery (“validation theater”)
These tools are useful for cheap, early learning but must be followed by real customer contact.
5. Empower “visionary builders” who can analyze, build, and sell
De Vries argues that value is shifting toward individuals who combine:
- Analyze (research, discovery)
- Build (prototypes, tools, processes)
- Sell (transfer conviction to stakeholders)
AI lets “dreamers” also become “doers,” so innovators should cultivate this end‑to‑end, builder‑seller profile rather than staying only in the analyst role.
6. Provide a sanctioned AI sandbox to avoid shadow IT
He notes that when innovators can suddenly build very polished prototypes with AI, it can:
- Create pressure on IT (“Why can’t you ship this tomorrow?”)
- Lead to ungoverned, insecure shadow IT if internal tools lag behind external ones
His advice is to offer an official, governed sandbox with reasonably current AI capabilities so people don’t feel forced to go rogue.
7. Expect the prototype to production transition to remain the hardest part
De Vries emphasizes that moving from working prototype/MVP to scaled production is still the key bottleneck:
- AI can help with business cases, narratives, and ROI modeling
- But you still face organizational constraints: backlogs, governance, politics, resource competition
Innovators should plan for this “innovation chasm” explicitly, rather than assuming AI alone will magically eliminate it.
(Featured image by Alexander Nrjwolf on Unsplash.)
